Efficacy, Safety and Pharmacokinetics of Once-Daily Saquinavir Soft-Gelatin Capsule/Ritonavir in Antiretroviral-Naive, HIV-Infected Patients

Context Once-daily HIV treatment regimens are being used in clinical practice with the objective of improving patient acceptance and adherence. Objective To evaluate the efficacy and safety of saquinavir-soft-gelatin capsule (SGC)/ritonavir combination (1600 mg/100 mg) vs efavirenz (600 mg) both once daily and combined with 2 nucleoside analogs twice daily. Setting Twenty-six centers in the United States, Canada, and Puerto Rico. Patients A total of 171 antiretroviral naive HIV-infected individuals were enrolled in a 48-week, phase 3, open-label, randomized study. Main Outcome Measure Proportion of patients with HIV-RNA levels < 50 copies/mL. The pharmacokinetic profile of saquinavir-SGC was analyzed in a subset of randomly selected patients. Results In the primary intent-to-treat population at week 48, 51% (38/75) and 71% (55/77) of patients in the saquinavir-SGC/ritonavir and efavirenz groups, respectively, achieved HIV-RNA suppression < 50 copies/mL (P = .5392, 95% 1-sided confidence interval [CI] = -33.5%). In the on-treatment (OT) population, 73% (38/52) and 93% (54/58) of patients in the saquinavir-SGC/ritonavir and efavirenz groups, respectively, had effective viral suppression < 50 copies/mL (P = .5015, 95% 1-sided CI = -33.4%). Mean CD4+ cell counts increased by 239 and 204 cells/microliters (mcL), in the saquinavir-SGC/ritonavir and efavirenz groups, respectively, in the OT analysis (P = .058). Both regimens were reasonably well tolerated, although more gastrointestinal adverse events were reported with saquinavir-SGC/ritonavir. Pharmacokinetic profiles in 6 patients showed an observed median Cmin at 24 hours of 429 ng/mL (range, 681750 ng/mL). Conclusion Once-daily efavirenz was statistically superior to once-daily saquinavir-SGC/ritonavir. Gastrointestinal adverse effects were commonly associated with treatment failure in the saquinavir-SGC/ritonavir arm of the study.


Introduction
The introduction of highly active antiretroviral therapy (HAART) has produced dramatic reductions in morbidity and mortality rates associated with HIV-1 infection in the United States. [1,2] In clinical trials, HAART has reduced plasma HIV-1 RNA levels to less than 400 copies/mL in 60% to 90% of patients. [3] Strict adherence to HAART is necessary to prevent viral replication and the emergence of drug-resistant viruses, which can compromise the final therapeutic benefit. [4,5] Treatment regimens with improved dosing schedules, such as once-daily dosing, are likely to improve patient acceptance and adherence.
[6] Furthermore, once-daily dosing of HAART may be particularly beneficial in the implementation of directly observed therapy (DOT) in prisons, at needle-exchange sites, and in drug rehabilitation programs. [4] To date, 6 antiretroviral (ARV) agents, efavirenz, tenofovir, didanosine, lamivudine, coformulated lamivudine/ abacavir, atazanavir and amprenavir boosted with ritonavir are approved for once-daily dosing by the US Food and Drug Administration (FDA). However, in addition to these drugs, there are several agents such as nevirapine and other boosted protease inhibitors (PIs) that are being used once daily in clinical practice based on their half lives. The availability of a wide choice of once-daily treatment regimens has made it easier for HIV-infected patients to find an optimal therapy that suits their lifestyle. While efavirenz appears to be an ideal once-daily treatment option due to its potency and convenience with a low pill burden, [7] adverse events, in particular relating to the central nervous system, have been reported following administration,[8] which may potentially limit its use in some HIV-infected patients. Additionally, efavirenz may not be appropriate in some settings because it may have teratogenic effects. [8] Ritonavir-boosted PI regimens are widely used in clinical practice, [9,10] because several boosted PI combinations have pharmacokinetic profiles that support once-daily dosing. [11] These include saquinavir/ritonavir, [12] amprenavir/ritonavir, [13] fosamprenavir/ritonavir, [14] lopinavir/ritonavir [15] and atazanavir/ritonavir. [16] In initial studies, positive pharmacokinetic and efficacy data have been observed with the use of once-daily saquinavir/ ritonavir in PI-naive and experienced individuals. [12,17] Therefore, to further investigate the saquinavir/ritonavir boosted PI combination as a potential once-daily treatment regimen, we evaluated the efficacy and safety of saquinavir-soft-gelatin capsule (SGC)/ritonavir 1600 mg/ 100 mg vs efavirenz 600 mg in a prospective, randomized, multicenter clinical trial. Both treatment regimens were administered once daily in addition to 2 nucleoside reverse transcriptase inhibitors (NRTIs) (twice daily) as part of combination HAART therapy regimens to ARV-naive, HIV-infected individuals. As part of this clinical study, the pharmacokinetic profile of saquinavir-SGC was assessed in a subset of patients.

Study Design
This was a phase 3, open-label, randomized, multicenter study conducted at 26 centers in the United States, Canada, and Puerto Rico. Antiretroviral-naive, HIV-infected adults were randomized to receive either saquinavir-SGC/ ritonavir (1600 mg/100 mg, 9 pills) or efavirenz (600 mg, 3 pills) once daily, both in combination with 2 NRTIs twice daily. An interim analysis was planned at week 24, with a follow-up extension to week 48. Once patients reached 48 weeks, they had the option of continuing in the study until a common study closure (CSC) date or of withdrawing at that time. The CSC date was the date on which the last randomized patient reached 48 weeks of treatment.
The NRTIs allowed in the study included the following agents: zidovudine (AZT, Retrovir), lamivudine (3TC, Epivir), 3TC/AZT (Combivir), stavudine (d4T, Zerit), or didanosine (ddI, Videx). The NRTIs were selected by the investigators following consultation with the study participants, with the exception of didanosine, which was dosed twice daily according to the FDA approval status at the time this study was conducted. Patients were randomized via stratification of the screening plasma HIV-RNA value (500075,000 copies/mL vs > 75,000 copies/ mL). Saquinavir-SGC and ritonavir were to be taken at the same time and with food. The study was designed and monitored in accordance with Good Clinical Practices. Prior approval was obtained from institutional review boards/local ethics committees of the participating centers and written informed consent was obtained from all study participants.

Patients
The defined study population consisted of male or female (nonpregnant, nonnursing) adults (18 years or older), with HIV-RNA values  5000 copies/mL and CD4+ cell counts  75 cells/mcL. All patients were ARV-naive (no more than 2 weeks of previous ARV therapy since HIV diagnosis). Patients with significant laboratory abnormalities, active hepatitis B or hepatitis C, severe hepatic impairment, or malignancy requiring chemotherapy or radiation therapy were excluded from the study.

Study Populations
All efficacy analyses were performed on the primary intent-to-treat (ITT) and on-treatment (OT) populations. The primary ITT population included all patients who received at least 1 dose of study drug after randomization and had at least 1 efficacy evaluation. The OT population included those patients in the study who had an observational value for all efficacy variables and laboratory measurements, at that particular timepoint in the study. The safety-evaluable population included all randomized patients who received at least 1 dose of study medication and had at least 1 postbaseline safety assessment.

Efficacy and Safety Measures
The primary efficacy variable was the proportion of patients with plasma HIV-RNA values < 50 copies/mL using the Ultra-Sensitive assay (Roche Diagnostics), at weeks 24 and 48. Secondary efficacy analyses included the change from baseline in CD4+ cell counts. Primary and secondary variables were assessed at baseline (week 0), every 4 weeks until week 24, and every 8 weeks until week 48, or at study discontinuation.
Adverse events were graded by intensity, and their relationship to the study medications was assessed. Fasting lipid profile was also assessed. Laboratory abnormalities were graded by severity. Marked laboratory shifts were defined as a  3 grade shift from baseline (grade 03 or 4 and grade 14).

Pharmacokinetic Subanalysis
A subanalysis was performed to assess the pharmacokinetic profile of saquinavir-SGC in a subset of randomly selected patients. Intensive pharmacokinetic assessments were performed on patients from 2 preselected study centers after the completion of 4 weeks of therapy. Following a standard breakfast, saquinavir-SGC/ritonavir 1600 mg/ 100 mg was administered with 2 NRTIs. Samples for analysis of saquinavir-SGC (7 mL) were taken predosing (hour 0) and at 1, 2, 3, 4, 6, 8, 12, and 24 hours post-dose. At week 4, single trough plasma concentrations of saquinavir-SGC were also determined in a larger group of patients. For patients who took their dose in the morning, single samples were taken before their next morning dose (approximately 24 hours later) and for those who took their dose in the evening, samples were taken at least 12 hours after the previous dose.

Pharmacokinetic Analysis
Saquinavir-SGC plasma concentration was assayed using a validated SCIEX API liquid chromatography extraction method with a mass spectrophotometric (LC-MS) detection system. The method was validated for a range of 53000 ng/mL. For patients who underwent intensive pharmacokinetic evaluation, the variables assessed were area under the curve over 24 hours (AUC 024 h ), maximum plasma concentration (C max ) and observed trough plasma concentration (C min ). C max and C min were read directly from the observed data. The AUC data were obtained by noncompartmental analysis.
In patients with a single trough plasma sample, C min at 24 hours (C 24 h ) was the pharmacokinetic variable of interest and saquinavir-SGC C 24 h was mathematically extrapolated to reflect 24-hour levels, assuming an elimination half-life of 56 hours. [18] The equation C 1 = C 0 e(-k el • t) was used to calculate the saquinavir-SGC C 24 h extrapolated levels where: C 1 = concentration extrapolated at 24 hours; C 0 = observed concentration; k el = the elimination rate constant for the drug (k el = 0.693/t 1/2); t = difference in time between C 1 and C 0 ; and t 1/2 is defined as the half-life of the drug.
Descriptive statistics for the pharmacokinetic parameters of saquinavir-SGC were summarized to compare them with historical values obtained with the licensed dose of saquinavir-SGC (1200 mg 3 times daily [TID]).

Statistical Analyses
This study was designed to compare the efficacy of oncedaily dosing of saquinavir-SGC/ritonavir 1600 mg/100 mg vs efavirenz 600 mg at weeks 24 and 48, using the proportion of patients with HIV-RNA values < 50 copies/mL (response rate). Group differences in the proportion of patients with HIV-RNA levels < 50 copies/mL were analyzed using a noninferiority test. The Farrington-Manning method for testing noninferiority was used to derive P values and to construct 1-sided 95% confidence intervals (CI) of the rate difference. [19] If the limit of the 95% 1sided CI of the difference rate was greater than -20%, then noninferiority was concluded. Mean changes from baseline in CD4+ cell counts were analyzed using an analysis of variance (ANOVA) model. Safety variables were summarized using descriptive statistics for the entire study period up to the CSC date. Fisher's exact test was used to statistically compare the incidence of adverse events in both treatment groups.

Patients and Study Course
From July 1999 to October 2001, 269 patients were screened and 171 were randomized to treatment ( Figure  1). Six patients withdrew from the primary ITT population before receiving study medication, giving a total of 165 patients who received study drug. Of these, 13 patients were excluded because no postrandomization on-drug efficacy evaluation was available. The resulting primary ITT population included 75 patients in the saquinavir-SGC/ritonavir group and 77 patients in the efavirenz group. Of the primary ITT population, 108 patients completed all 48 weeks of the study (51 in the saquinavir-SGC/ritonavir group and 57 in the efavirenz group). At week 48, the OT population consisted of 110 patients: 52 and 58 patients in the saquinavir-SGC/ritonavir and efavirenz arms, respectively. The safety population consisted of 161 randomized patients: 81 and 80 patients in the saquinavir-SGC/ritonavir and efavirenz group, respectively.
At the CSC date, 51/152 (34%) patients had withdrawn from the primary ITT population: 28/75 (37%) patients from the saquinavir-SGC/ritonavir group and 23/77 (30%) patients from the efavirenz group ( Figure 1). Of these, 18 patients withdrew due to failure to return for reassessment (8 in the saquinavir-SGC/ritonavir arm and 10 in the efavirenz arm), 13 patients refused treatment/ withdrew consent (11 in the saquinavir-SGC/ritonavir arm and 2 in the efavirenz arm), and 12 withdrew due to adverse events (8 in the saquinavir-SGC/ritonavir arm and 4 in the efavirenz arm). The demographic and baseline characteristics of the patients in the saquinavir-SGC/ritonavir and efavirenz arms were similar, with respect to gender, age, race, baseline HIV-RNA levels and CD4+ counts ( Table 1). The majority (73%) of patients in both arms was male, 47% were black, and the age range was 1961 years (median, 36 years).

Changes in CD4+ Cell Counts
At week 24, the mean increase in CD4+ cell counts was 170 cells/mcL in the saquinavir-SGC/ritonavir arm compared with 143 cells/mcL in the efavirenz arm (P = .344; OT population). At the 48-week study endpoint, this had increased to 239 cells/mcL and 204 cells/mcL, in the saquinavir-SGC/ritonavir and efavirenz groups, respectively (P = .058; OT population). Completed 48-week study n = 51

Study design and completion status
Completed 48-week study n = 57

Safety
Overall, adverse events were reported by 98% and 94% of patients in the saquinavir-SGC/ritonavir and efavirenz groups, respectively. The majority of adverse events were mild or moderate in intensity. During the entire study period, 11% (18/161) of patients in the safety population discontinued treatment due to adverse events. In the saquinavir-SGC/ritonavir group 15% (12/81) of patients experienced adverse events that led to withdrawal and 8% (6/80) of patients withdrew from the efavirenz group due to adverse events. The most commonly reported adverse events leading to withdrawal in the saquinavir-SGC/ritonavir group were gastrointestinal in nature, including nausea and vomiting. In contrast, weight loss, peripheral neuropathy and abnormal dreams were the reported adverse events leading to withdrawal in the efavirenz group.
Adverse events of moderate, severe or life-threatening intensity that were considered possibly or probably related to study medication occurring in 2% or more of patients are shown in Table 2 . Overall, 42% (34/81) and 29% (23/80) of patients in the saquinavir-SGC/ritonavir and efavirenz groups, respectively, reported at least 1 adverse event of moderate, severe, or life-threatening intensity considered possibly or probably related to study medication. Among adverse events, the rates of nausea differed significantly among treatment groups (Table 2).
In total, 15 patients (7 [8.6%] and 8 [10.0%] patients in the saquinavir-SGC/ritonavir and efavirenz groups, respectively) reported at least 1 serious adverse event, none of which were considered to be related to study medication. One patient in the saquinavir-SGC/ritonavir group died due to myocardial infarction, which again was not considered to be related to study medication.

Laboratory Variables
Overall, laboratory grade shifts in hematology and chemistry values were similar in both treatment groups. Grade 3/4 total cholesterol laboratory shifts were reported in 3 patients in the saquinavir-SGC/ritonavir group compared with 7 patients in the efavirenz group. Mean triglyceride and total cholesterol levels assessed throughout the 48week study period are presented in Figure 3. At weeks 24 and 48, mean changes in triglyceride values from baseline were 29.4 mg/dL and 51.2 mg/dL, respectively, in the saquinavir-SGC/ritonavir group, compared with 27.0 mg/ dL and 71.4 mg/dL, respectively, in the efavirenz group. The differences between groups were not statistically significant at weeks 24 and 48. Mean changes in total cholesterol levels were comparable between the 2 treatment arms throughout the 48-week study period. Mean changes in high-density lipoprotein cholesterol at weeks 24 and 48 were 7.9 mg/dL and 8.7 mg/dL, respectively, in the saquinavir-SGC/ritonavir group compared with 9.3 mg/ dL and 12.1 mg/dL, respectively, in the efavirenz group.

Pharmacokinetics
Of 72 patients who were enrolled in the pharmacokinetic sub-analysis, 6 underwent intensive pharmacokinetic evaluation. Sixty-five patients had single plasma samples taken at week 4 and data from 44 patients were suitable for evaluation of trough concentrations (data from the remaining 21 could not be extrapolated because the samples from 11 patients were taken less than 12 hours after the previous dose and the remaining 10 patients were noncompliant). Furthermore, of the 6 patients included in the intensive pharmacokinetic analysis, 1 patient was noncompliant and, therefore, not included in the singletrough evaluable cohort of patients. The single-trough evaluable cohort, therefore, consisted of 49 patients (5 patients with intensive pharmacokinetics and 44 with single pharmacokinetic evaluations).
Percentage of patients treated with either saquinavir-soft gelatin capsule/ritonavir or efavirenz both once-daily in combination with 2 nucleoside reverse transcriptase inhibitors, with undetectable HIV-RNA levels (< 50 copies/mL), as detected by the Ultra-Sensitive assay

Intensive Pharmacokinetic Analysis at 4 Weeks
The intensive pharmacokinetic profile of the 6 unselected patients evaluated at 9 time points over 24 hours is shown in Figure 4. The median observed C min at 24 hours, C max and AUC 024 for these 6 patients were 429 ng/mL (range, 681750 ng/mL), 6435 ng/mL (range, 172013,400 ng/ Mean levels of triglycerides and total cholesterol in the safety population of patients treated with either saquinavir-soft gelatin capsule/ritonavir or efavirenz both once daily in combination with 2 nucleoside reverse transcriptase inhibitors mL), and 66,920 ng.h/mL (range, 10,542137,563 ng.h/ mL), respectively.

Saquinavir-SGC Trough Levels at Week 4
The median saquinavir-SGC C 24 h extrapolated level at week 4 for all evaluable patients (n = 49) was 376 ng/mL (range, 2.55709 ng/mL). Eight patients had saquinavir-SGC C 24 h extrapolated levels under the in vivo concentration for half-maximal effect (EC 50 ) value of 50 ng/mL (range, 2.545.2 ng/mL). [20] Despite this, these patients had sustained HIV viral load suppression up to week 24. There was no significant correlation between saquinavir-SGC C min values and the reduction in plasma HIV RNA values from baseline to week 48.

Discussion
This 48-week study compared the efficacy and safety of the once-daily HIV treatment regimens, including saquinavir-SGC/ritonavir 1600 mg/100 mg, to that of efavirenz 600 mg. The primary ITT analysis showed that a higher proportion of patients in the efavirenz-treated group achieved HIV-RNA levels < 50 copies/mL compared with the saquinavir-SGC/ritonavir-treated group. More patients withdrew from randomized treatment due to gastrointestinal disorders in the saquinavir-SGC/ritonavir arm. The lower virologic response rate observed in the saquinavir-SGC/ritonavir arm in the ITT population was predominately due to the high rate of discontinuations, with 37% of patients withdrawing from the saquinavir/ ritonavir arm, compared with 30% from the efavirenz arm. These withdrawals were related to gastrointestinal side effects, and possibly higher pill counts (patients taking saquinavir-SGC/ritonavir had to take 9 capsules at 1 time compared with 3 efavirenz capsules, together with the background NRTIs). This study included a high percentage of African-American patients and no specific adherence intervention was provided to overcome the difference in pill burden in the 2 treatment arms.
A previous comparative randomized trial of the boosted PI saquinavir-SGC/ritonavir has also shown lower virologic response rates, which have been attributed to differences in study withdrawal rates as a result of gastrointestinal toxicities in the saquinavir-SGC study arms. [21] In the MaxCmin 2 study, which compared twice-daily saquinavir-SGC/ritonavir with lopinavir/ ritonavir, comparable viral suppression was demonstrated in the OT analyses; however, the difference in study withdrawal rates due to gastrointestinal toxicity in the saquinavir-SGC/ritonavir arm led to a lower overall response in the ITT analyses. [21] The gastrointestinal toxicities encountered in the saquinavir-SGC study arm of MaxCmin 2 were attributed to the glyceride capmul component of the SGC formulation of saquinavir.
Saquinavir is available in 2 formulations: the SGC formulation, Fortovase, and the original hard capsule (HC) saquinavir mesylate formulation, Invirase. In the present study, the SGC formulation of saquinavir was used. As a sole PI (unboosted), the oral bioavailability and effectiveness of saquinavir with the SGC formulation is higher than that of the HC formulation. This difference is due to the glyceride excipient component of the SGC formulation, capmul, which allows saquinavir to dissolve and disperse rapidly upon administration. However, the SGC formulation is associated with more gastrointestinal adverse events. [22,23] In a recent study of the saquinavir/ ritonavir boosted regimen, the HC formulation of saquinavir mesylate demonstrated similar exposure to the SGC formulation but was better tolerated. [22,24] Consequently, it might be expected that the tolerability issues encountered in this study could be ameliorated if the HC saquinavir mesylate formulation of saquinavir was used.
Interestingly, in a further study the boosted saquinavir-HC/ritonavir 1600 mg/100 mg regimen demonstrated potent viral suppression (89% of patients had HIV RNA values < 50 copies/mL) with no patients withdrawing from the study due to adverse events. [25] Of note, the saquinavir mesylate is now available as a 500-mg filmcoated tablet, which is similar in size to the saquinavir mesylate 200 mg HC.
Changes in lipid levels have been associated with the use of NRTIs and PIs and, consequently, the impact on lipid levels should be an important consideration when choosing a treatment regimen. [26] With PIs, this risk has been shown to be greatest with ritonavir, intermediate with amprenavir and nelfinavir, and lowest with indinavir and saquinavir. [27,28] The recently approved PI, atazanavir appears to have little effect on lipid concentrations. [16,27] The lower doses of ritonavir (100200 mg) that are suitable for boosted saquinavir/ritonavir regimens may result in some increases of fasting triglyceride and cholesterol levels. [29] The data from this study confirm a low lipid profile risk associated with a once-daily saquinavir-SGC/ritonavir dose of 1600 mg/100 mg. During this study, data evolved regarding the contribution of NRTIs to hyperlipidemia and a post-hoc analysis was conducted that demonstrated that the predicting factor for hyperlipidemia in this study appeared to be the use of stavudine as the background NRTI. [30] Efavirenz once daily is a highly effective and commonly prescribed part of ARV therapy; however, it may not be suitable for all patients. Patients taking efavirenz may experience central nervous system adverse events such as dizziness, insomnia, poor concentration, and vivid dreams.
[8] With the recent classification of efavirenz as pregnancy category D caution, it is advised in the treatment of women of child-bearing potential. Also, efavirenz use may be a cause for concern in drug users currently taking methadone because efavirenz-mediated reduction of methadone blood concentrations has also been reported. [31,32] No methadone dose adjustments have been found to be required in patients receiving once-daily saquinavir/ritonavir 1600 mg/100 mg therapy. [33] The results of the pharmacokinetic sub-analysis illustrated that a once-daily regimen including saquinavir-SGC/ ritonavir 1600 mg/100 mg provides a median AUC 024 h of 66,920 ng.h/mL. This is significantly higher than might be anticipated following the unboosted saquinavir SGC regimen of 1200 mg 3 times daily, where the AUC 08 h is 7249 ng.h/mL. [34] In addition, the median observed C min at 24 hours for saquinavir-SGC of 429 ng/mL was substantially higher than the EC 50 value of 50 ng/mL. [20] There is some evidence that failure on PI therapy can be linked to low plasma concentrations, [35] but in this study no relationship was found between HIV RNA concentrations and saquinavir C min values. [36] Currently, with antiretroviral therapy, it is unknown which pharmacokinetic parameter correlates best with efficacy. [37] While there is some evidence that failure on PI therapy can be linked to low C min plasma concentrations, [20] in this study no relationship was found between HIV RNA concentrations and saquinavir C min values.
In conclusion, once-daily efavirenz was statistically superior to the saquinavir-SGC/ritonavir (1600 mg/100 mg) once-daily regimen, as part of combination ARV therapy. Gastrointestinal adverse effects were commonly associated with (ITT) treatment failure in the saquinavir-SGC/ ritonavir arm of the study, which may be attributed to the SGC formulation used in this study. Future studies should evaluate the use of the saquinavir mesylate formulation, which has demonstrated similar exposure to the ritonavirboosted SGC formulation with a better tolerability profile and by virtue of the 500-mg tablet also provides a lower pill burden.

Funding Information
This study was supported by a research grant from Roche Pharmaceuticals.